Looptime Discussion:

At 100% efficiency, a 2300Kv motor spins at 36,800rpm at 16v wide open.

The motors we use on 5" have 12 poles, which means that the ESC is switching 12 times per revolution, or 441,600 time per minute. (36,800*12)
This mean that the ESCs are switching at 7,360 per second at 16v wide open.

How could it be beneficial to run a looptime (hz) that is any higher than the ESC is even switching and adjusting? I can see that 8Khz could make some difference since you're now computing at the same rate as the ESC is physically working, but I see talk of 16Khz and now 32Khz coming and I'm just not understanding how there could be any real benefit...


I have tried 4K, 8K and 32K, Simply cannot see any difference from 8khz and higher frequency rates. Even 4khz is hard to notice any difference.

Basically it has little difference but in analog signal increasing loop time just sends the signal multiple times per switch so if the analog signal is slightly off it will average out the signal kind of thing.

A bit of a tangent but I think the ESC switches the FET many times per pole - the length of time that it's switched on for is proportional to the throttle amount (it's a PWM signal going to each phase)

But just because the ESC output stage is switching faster doesn't mean that it's reading faster anyway. It's quite feasible to code it in a way that adjusts the throttle once per revolution, or once per pole, or many times per pole.

But I fly Kiss anyway so I guess I'm just happy with 1Khz

From what i read, the advantage of 32kHz is simply less delay. Downside is that noise becomes an even bigger issue.